Monitoring

The Monitoring tab, minus the browser. Same data, same scoping, same permissions, driven by your signed-in session.

octri auth login
octri projects use
octri monitoring summary

Is anything broken

octri monitoring summary
octri monitoring summary --range 7d

Last 24h
  events           1,284
  errors           37
  error rate       2.9%
  distinct issues  4

--range takes 1h, 6h, 24h, 7d, 30d or 90d and works on every windowed command here.

Issues

octri monitoring issues
octri monitoring issues --status unresolved --level error
octri monitoring issues --q checkout --limit 50

Each row carries a short id. Pass the full id from --json output, or the visible one, to read the issue:

octri monitoring issue <id>

That prints the stack, and marks which frames resolved against uploaded symbols. If none did, the release those errors came from has no symbols and the CLI says so.

Triage

octri monitoring resolve <id>
octri monitoring ignore <id>
octri monitoring reopen <id>
octri monitoring comment <id> "Fixed in 2.4.1, watching"

Comments and status changes appear in the dashboard against your name, the same as if you had clicked.

Raw events

Issues group by fingerprint. When you need every occurrence instead:

octri monitoring logs
octri monitoring logs --level error --q timeout --limit 100

Did that deploy make things worse

octri monitoring releases

One row per release, with its error rate, new issues, and regressions. This is the fastest way to answer the question after a deploy.

octri monitoring traces
octri monitoring traces <traceId>
octri monitoring performance

performance lists the slowest transactions with p50 and p95, then any suspected N+1 query patterns underneath.

Alerts

octri monitoring alerts
octri monitoring alerts create "Checkout errors" \
  --webhook https://hooks.example.com/octri \
  --threshold 25 --window 15
octri monitoring alerts evaluate
octri monitoring alerts delete <id>

--slack takes a Slack webhook instead. --kind accepts threshold, new_issue or regression, defaulting to threshold.

alerts evaluate runs every rule once, now, and reports which fired. Useful for checking a rule does what you think before you wait for it.

Synthetic checks

octri monitoring checks
octri monitoring checks generate
octri monitoring checks run <id>

checks generate derives one check per safe endpoint from your current spec. If the spec has no absolute server URL, pass one:

octri monitoring checks generate --base-url https://api.example.com --every 5

Turning it on

octri monitoring status
octri monitoring enable

After enabling, rebuild your SDKs so they carry the ingest configuration. A generated SDK reports to the endpoint that was baked in when it was built.

octri monitoring config

That prints MONITORING_URL, MONITORING_ENVIRONMENT and MONITORING_TOKEN for storing as CI secrets. The token stays hidden unless you pass --reveal-token.

Symbols

octri monitoring sourcemaps upload ./dist
octri monitoring sources upload ./src
octri monitoring sourcemaps list --release "$GIT_SHA"

Signed in, the connection is resolved from the selected project. In CI, pass --url, --token and --environment, or set them in the environment. Details in Source maps, Source files and CI setup.

Sending a test event

octri monitoring test-event

One synthetic error, delivered through the real ingest path. It shows up in octri monitoring issues a few seconds later, which confirms the connection end to end.